Vol. 6 No. 1 (2023)

Open Access

Article

Article ID: 578

A fast and accurate camera-IMU calibration method for localization system

by Xiaowen Tao, Pengxiang Meng, Bing Zhu, Jian Zhao

Insight - Automatic Control, Vol.6, No.1, 2023; 273 Views, 44 PDF Downloads

Autonomous driving has spurred the development of sensor fusion techniques, which combine data from multiple sensors to improve system performance. In particular, a localization system based on sensor fusion, such as Visual Simultaneous Localization and Mapping (VSLAM), plays a crucial role in environment perception and serves as the foundation for decision-making and motion control in intelligent vehicles. The accuracy of extrinsic calibration parameters between the camera and IMU is of utmost importance for precise positioning in VSLAM systems. However, existing calibration methods are often time-consuming, rely on complex optimization techniques, and are sensitive to noise and outliers, leading to potential degradation in system performance. To address these challenges, this paper presents a fast and accurate camera-IMU calibration method based on space coordinate transformation constraints and SVD (Singular Value Decomposition) tricks. The method involves constructing constraint equations by ensuring the equality of rotation and transformation matrices between camera frames and IMU coordinates at different time instances. Subsequently, the external parameters of the camera-IMU system are solved using quaternion transformation and SVD techniques. To validate the proposed method, experiments were conducted using the ROS (Robot Operating System) platform, where camera images and velocity, acceleration, and angular velocity data from the IMU were recorded in a ROS bag file. The results demonstrate that the proposed method achieves reliable camera-IMU calibration parameters, requiring less tuning time and exhibiting reduced uncertainty.

Open Access

Article

Article ID: 569

Speed control of PMBLDC motor drive powered by solar PV array using P, PI, and PID controllers: A comparison study

by Satish Kumar Doniparthi, S. B. Ron Carter, Amit Vilas Sant, Ali Moghassemi

Insight - Automatic Control, Vol.6, No.1, 2023; 780 Views, 28 PDF Downloads

Because of their high efficiency, better starting torque, and minimal electrical noise, permanent magnet brushless DC (PMBLDC) motors are frequently used in a variety of industrial applications. The speed of PMBLDC motors is controlled by a variety of controllers. In this study, P, PI, and PID controllers are used to compare the speed control of a permanent magnet brushless DC motor drive powered by solar PV arrays. The Perturb & Observe (P&O) technique is used to find the MPPT. The drive system’s simulation results for various operation modes, such as constant and variable load circumstances, are examined and evaluated. When using a PID controller instead of a P or PI controller, the drive performs better at controlling speed. The MATLAB/Simulink software was used to model, control, and simulate the permanent magnet brushless DC motor drive. The whole drive system is put into operation with the help of the dSPACE MicroLab Box 1202.

Open Access

Article

Article ID: 579

Control charts for processes with variable mean

by Mohammad Saber Fallahnezhad, Farideh Sadeghi, Amir Ghalichehbaf

Insight - Automatic Control, Vol.6, No.1, 2023; 57 Views, 15 PDF Downloads

Despite of strong ability and performance of control charts to control and monitor processes, they have some problems in practical applications. If control chart’s limits are not properly designed then we receive false alarms. For example, several observations may be outside the control limits when the mean of process is in-control. Not considering the variation of the process mean at each sampling time may lead to this error. The process may be adjusted at specific mean but different working conditions and different operators may change mean of the process and it may have a small deviation from its predetermined value and this problem can lead to wrong implementation of control charts. In this paper, the effects of variable mean on control charts are analyzed. It is assumed that the mean of observation varies over time but its probability distribution is normal probability distribution function. It is observed that long-term process mean control chart generates false alarms.